Metric Definition Template (Owner, Formula, Action Thresholds)

Metric Definition Template

Use this template to capture a complete, unambiguous definition for every metric you rely on. A well-defined metric ties measurement to a clear purpose and named owner, explains calculation and data sources, sets targets and thresholds, and spells out actions when thresholds are crossed.

Template fields (fill every field)

1. Name

   Short, descriptive name used in reports (avoid internal jargon). Example: "On-Time Delivery (%)".

2. Purpose / Hunger Served

   One or two sentences: why this metric exists and which decision or improvement it supports. Link to the business question it helps answer.

3. Owner (primary)

   Role or person responsible for the metric, data quality, and triggering actions. Include a backup owner and contact info or team.

4. Stakeholders

   Who uses this metric and who must be notified when thresholds are breached.

5. Definition / Calculation

   Exact formula with names for numerator and denominator. Define inclusion/exclusion rules, time windows, and any filters. Provide a worked numeric example.

   Example format: Calculation = (Numerator) ÷ (Denominator) × 100. Numerator: count of orders delivered on or before committed date. Denominator: total orders scheduled for delivery in the period. Exclude: test orders, cancellations.

6. Data source & extraction details

   System(s) and fields used, query name or path, data owner, update frequency of the source, and any transformation steps (ETL). Note known data quality issues.

7. Unit / Type

   Percent, count, dollars, days, ratio, index, yes/no. Indicate if it's a leading or lagging indicator.

8. Frequency & granularity

   How often the metric is calculated and reported (e.g., daily, weekly, monthly) and the level of aggregation (site, product line, customer segment).

9. Baseline & target

   Baseline value (with date), short- and long-term target, rationale for the target, and how targets are reviewed.

10. Action thresholds & required responses

    Define named threshold bands and the specific actions, owners, and escalation path for each. Be precise about when an action should trigger.

    • Green (OK) — numeric range; monitoring frequency; no immediate action required.
    • Amber (Watch) — numeric range; owner to investigate within X business hours/days, run root-cause checklist, schedule corrective action plan.
    • Red (Action) — numeric range; immediate escalation to named role, convene huddle within Y hours, implement predefined countermeasures.
11. Visualization & recommended dashboard view

    Suggested chart type (trend, control chart, stacked bar), recommended time window, related filters, and KPIs to display alongside this metric for context.

12. Interpretation guidance

    Short notes on how to interpret common fluctuations, seasonality, anomalies, and sample-size considerations. Warn about situations where the metric may be misleading.

13. Related metrics / dependencies

    List upstream/downstream metrics and processes that influence this metric. Note any known causal relationships or confounders.

14. Controls against gaming

    Practical steps and monitoring to reduce incentive-driven distortions (e.g., audit samples, cross-check metrics, rotate definitions).

15. Review cadence & version history

    Next review date, review owner, and a brief change log (date, change, author, reason).

16. Notes

    Any additional context, links to reference documentation, calculation scripts, or queries.

Example (completed)

Common pitfalls & quick tips

• Don't leave the calculation vague — spell out numerator, denominator, time window, and exclusions.
• Avoid mixing units or combining unrelated events in one metric (e.g., mixing returns and new sales in a single count).
• Prefer simple, observable measures over opaque indexes unless the index is well documented and justified.
• Be explicit about the metric's role: leading (predictive) or lagging (outcome). Use a mix of both to guide action.
• Document data quality issues and how they are handled (nulls, duplicates, late arrivals).

When to use this template

Use for any metric used in decision-making, performance dashboards, incentive programs, or operational escalation. Keep the filled template with the metric's dashboard and make it accessible to stakeholders.

Where this template helps most

Reduces ambiguity, prevents misinterpretation, strengthens accountability, and makes metric-driven actions faster and more consistent.